Decentralized coordination of autonomous swarms inspired by chaotic behavior of ants

In this paper, we propose a decentralized coordination algorithm for a group of mobile nodes, called an autonomous swarm, on a finite two-dimensional space, which can efficiently coordinate cooperatively the autonomous swarm to the optimal solution. Our algorithm is inspired by chaotic behavior of a single ant and self-organization behavior of the whole ant colony. To construct this algorithm, we firstly assume that each agent is a nonlinear oscillator presenting the chaotic behavior of a single ant. Then we establish a self-organization mechanism according to the self-organization behavior of the whole ant colony. Moreover, we analyze the convergence of the proposed algorithm. Finally, we experimentally evaluate the performance of our algorithm with the clustering and dispersion operations of a swarm. Comparison results of the proposed algorithm and the gradient-type one are also presented to illustrate the effectiveness of the proposed scheme in approximately global optimization for swarms.